© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 67983
    Prev
    Next
    INDEX
    Explanations

    thesis statement

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron fires most strongly on standalone, title‐style phrases or section headings made up of capitalized words (e.g. headings, product or article titles).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     zoude
    -1.55
    ครื่อง
    -1.52
    
    -1.50
     organisme
    -1.48
     pérd
    -1.46
     inay
    -1.45
     排骨
    -1.45
     visse
    -1.44
     poë
    -1.42
     horloge
    -1.41
    POSITIVE LOGITS
     What
    2.23
     How
    2.13
     Where
    1.95
     what
    1.95
     there
    1.90
     how
    1.88
     where
    1.81
     Don
    1.79
     There
    1.70
     Why
    1.70
    Activations Density 0.001%

    No Known Activations